Atypical Antipsychotics and the Risk of Diabetes in an Elderly Population in Long-Term Care: A Retrospective Nursing Home Chart Review Study

Transcription

1 Atypical Antipsychotics and the Risk of Diabetes in an Elderly Population in Long-Term Care: A Retrospective Nursing Home Chart Review Study Stewart G. Albert, MD, George T. Grossberg, MD, Papan J. Thaipisuttikul, MD, Jarod Scouby, BS, and Erin Green Objectives: Although atypical antipsychotics (AA) are associated with weight gain and diabetes mellitus (DM) in younger patients, it is not known whether these drugs would have a detrimental effect on weight gain or diabetes in a long-term care elderly population. Design: Retrospective chart review. Setting: Two community nursing homes (NH). Participants: Charts of 1678 subjects admitted between 2000 and 2006 were screened; data from subjects on AA were analyzed. Measurements: DM was defined by diagnosis in the Minimum Data Set (MDS), the prescription of diabetes medications, fasting blood glucose (BG) 126 mg/dl or higher, or random BG 200 mg/dl or higher. Worsening of DM was defined as beginning a medication in those treated with diet alone, or adding an additional medication to those already on antiglycemic therapy. Results: There were 154 subjects on AA, mean age (SD) years. Although there was no difference in age between the 101 women and the 53 men, there was a greater proportion of women 85 years or older compared with men (57% versus 40%, P.04). Alzheimer s disease (AD) was diagnosed in 17% and non-ad dementia in 34%. Weight was normal (BMI less than 25 kg/m 2 ) in 63%, overweight (BMI higher than 25 but less than 30 kg/m 2 ) in 29%, and obese (BMI higher than 30 kg/m 2 ) in 8%. Subjects were followed on AA for a median of 13.1 weeks (interquartile range ). Despite these medications, 32% lost more than 5% of body weight. DM was an admitting diagnosis in 21%. There were 4 new and 5 worsening cases of DM during their stay in the nursing home; however, of these 9 cases, 4 occurred before the institution of AA. There was no increased frequency of weight gain or DM among the various atypical agents. Conclusions: In an elderly NH population, there was no evidence that short-term use (median 13.1 weeks) of atypical antipsychotic agents was associated with the onset or worsening of DM. (J Am Med Dir Assoc 2009; 10: ) Keywords: Atypical antipsychotic agents; diabetes mellitus; nursing home; elderly Department of Internal Medicine, Division of Endocrinology, Saint Louis University School of Medicine, St. Louis, MO (S.G.A.); Department of Neurology and Psychiatry, Division of Geriatric Psychiatry, Saint Louis University School of Medicine, St. Louis, MO (G.T.G., P.J.T.). The authors have no conflicts of interest relating to this article. Some data from this study were presented as part of a Plenary Symposium titled Psychosis in the Elderly at the Silver Congress of the International Psychogeriatric Association, October 15, 2007, Osaka, Japan. Address correspondence to Stewart G. Albert, MD, Saint Louis University School of Medicine, 1402 South Grand Boulevard, St. Louis, MO albertsg@slu.edu Copyright 2009 American Medical Directors Association DOI: /j.jamda Atypical antipsychotic agents are the mainstay of therapy for patients with schizophrenia and the best available therapy for patients with dementia who have psychosis, aggression, agitation, and behavioral disturbances. 1 4 They improve quality of life in the patient and caregivers. Whereas these agents are less likely than typical antipsychotic agents to have side effects such as anticholinergicity, parkinsonism, and tardive dyskinesia, in younger subjects with schizophrenia they are associated with weight gain, glucose intolerance, and new onset of diabetes mellitus. 5 9 The atypical antipsychotic drugs all have in common a focused D (2)-dopaminergic antagonism, but each has different neuroreceptor binding to other neurotransmitters such as histamine, serotonin, and acetyl choline. 9 The alternate neuroreceptor binding may account for differing potentials for weight gain, 7 9 islet cell dysfunction, 10 and induction of diabetes. 11 The National Institutes of Health (NIH)-sponsored Clinical Antipsychotic Trials of Intervention Effectiveness-Alzheimer s Disease (CATIE-AD) Study 12 was designed to evaluate ORIGINAL STUDIES Albert et al 115

2 effectiveness and side effects of the outpatient use of atypical antipsychotics in the elderly. There was short-term improvement in clinical efficacy, which was later tempered by discontinuance due to drug intolerance. Weight gain of 0.4 to 1.0 pounds per month and cumulative weight gain of more than 7% was observed in 6% to 11% of the patients. The incidence of new-onset diabetes mellitus has not yet been reported in the CATIE-AD study. Whereas these changes may have clinical relevance in younger patients on these agents, these elderly people studied started at a body mass index (BMI) of kg/m 2, which is the upper limits of normal for BMI. Weight gain may not be detrimental in elderly individuals, especially those with dementing disorders such as Alzheimer s dementia, who may require medication to stimulate appetite and weight gain. Although the CATIE-AD study was a well-designed randomized trial of these agents in outpatients, there are many differences relative to the use of atypical antipsychotic agents in the nursing home population. First, patients in nursing homes are older and often have more severe behavioral problems. They may have disorders of appetite with attendant weight loss. They usually require longer-term therapy with atypical antipsychotic agents, rather than the median duration of 5.3 to 8.1 weeks in the CATIE-AD study. 12 We therefore carried out a retrospective chart review of patients who were residents of nursing homes taking atypical antipsychotic agents for clinically necessary therapy, to evaluate for changes in weight, and onset of diabetes mellitus. METHODS Subjects data were retrieved from 2 community nursing homes consulted by the university s division of geriatric psychiatry. The characteristics of the nursing homes were as follows: Nursing Home A (300 patient beds) and Nursing Home B (240 patient beds). Both are private, for-profit facilities. All charts of patients who were in these facilities between January 1, 2000, and December 31, 2006, were screened. No identifiers were obtained. Those who were on atypical antipsychotic medications were included for further analysis. Diabetes was defined as a diagnosis per the Minimum Data Set (MDS), the prescription of antiglycemic drugs (sulfonylureas, metformin, glitazones, glucosidase inhibitors, or insulin), 2 determinations of fasting blood glucose 126 mg/dl or higher or random blood glucose 200 mg/dl or higher. Data were analyzed for preexisting diabetes and for the new onset or worsening of diabetes mellitus. The latter was defined as beginning a medication in those treated with diet alone, or adding an additional medication to those already on antiglycemic therapy. Data were analyzed by analysis of variance, and if significant, then by post hoc analysis using the methods of Bonferroni using known predictors of diabetes, which included age, sex, body weight (and body mass index), and change in body weight. New-onset or worsening diabetes was analyzed by life table analysis (Kaplan Meier analysis 13 ). Statistics of proportions were determined by Chi square analysis and Fisher s exact test, as appropriate. Nonparametric data were analyzed by the Mann Whitney U test. Statistical procedures were performed with the statistical package Statistica for Windows (version 7, Statsoft, Inc, Tulsa, OK). Significance was defined as P.05 by 2-tailed testing. Data are expressed as mean SD or median with interquartile range unless otherwise specified. The study was approved by the Institutional Review Board of Saint Louis University. RESULTS There were 1678 charts screened, of which 154 subjects were identified to be on atypical antipsychotic agents (101 women and 53 men). The mean age was years. Although there was no difference in the mean age of women compared with men ( versus years, P.19, nonsignificant) (Table 1), there were more women older than 85 years (n 58, 57%) compared with men (n 21, 40%), P.04. As expected, the height and weight of the women were less than the values in men, but there were no differences in the mean BMI between women and men, which was the upper limit of normal kg/m 2 compared with kg/m 2 for women and men respectively (Table 1). Sixty-three percent were normal weight (BMI 25 kg/m 2 ), 29% were overweight (BMI kg/m 2 ), and 8% were obese (BMI 30 kg/m 2 ). Dementia was a diagnosis on admission in 79 subjects of the 154 subjects (51%). Alzheimer s dementia was documented in 26 (17%) and non-alzheimer s dementia in 53 (34%) of subjects. There were no differences in the frequency of Alzheimer s or non-alzheimer s dementia between men and women. Thirty-seven percent of subjects (n 57) were taking antidementia medication, of which 25 were instituted upon admission or earlier. Anxiety disorders were reported in 13% and depression in 38% of all subjects. All subjects were on atypical antipsychotic medications as per the inclusion criteria. The atypical antipsychotic medications were started months after admission to the nursing home. Quetiapine was used predominantly, followed in frequency by risperidone, olanzapine, aripiprazole, and ziprasidone. There was a greater frequency of use of quetiapine in men (60%) than in women (33%) when compared with the other available antipsychotic medications (P.001). The doses of all atypical antipsychotic agents were considered to be in the low dose range (Table 1). Despite being on these medications, only 20 (13% of subjects) gained more than 5% of body weight; whereas, 49 (32% of subjects) lost more than 5% body weight (Table 1). There was no difference in the frequency of weight change of either weight gain ( 5%) or weight loss ( 5 %) between men and women (Table 1). Among the various atypical antipsychotic medications there was not a greater frequency of weight gain for those on olanzapine (2 of 27) when compared to the proportion with weight gain on quetiapine (7 of 64) or risperidone (10 of 52) (chi square, not significant). Diabetes mellitus was an admitting diagnosis in 33 (21%) of the 154 subjects. Men compared with women, were more likely to have diabetes (36% versus 14%, P.002). Those with diabetes on admission were more overweight compared with those without diabetes (BMI versus Albert et al JAMDA February 2009

3 Table 1. Demographics of Subjects on Admission to Nursing Home All (154) Female (101) Male (53) P (Male vs. Female) Age, y 82.8 (8.0) 83.4 (8.2) 81.7 (7.5) NS Height, cm ( 10.5) [142] (7.0) [90] ( 9.7) [52].001 weight, kg 64.4 ( 16.0) [153] 60.9 ( 14.8) [100] 71.1 ( 16.0) [53].001 BMI kg/m ( 8.2) [142] 25.0 ( 9.8) [90] 23.6 ( 3.8) [53] NS Duration in NH, months, median 5.3 ( ) 6.5 ( ) 3.3 ( ) NS (interquartile range) Diagnosis on admission, no. of subjects Diabetes * Alzheimer s disease, no NS Non-Alzheimer s dementia, no NS Anxiety disorder NS Depression NS Atypical antipsychotic medications Start drug from admission, months 6.0 ( 15.7) 6.6 ( 16.6) 5.0 ( 13.7) NS Duration of drug maintenance, months Median (interquartile range) ( ) ( ) ( ) NS Drug, no. of subjects P.01 Quetiapine Risperidone Olanzapine Aripiprazole Ziprasidone Dose, mg/day Quetiapine 59.3 ( 56.4) [64] 51.2 ( 52.5) [32] 67.7 ( 59.7) [32] NS Risperidone 0.83 ( 0.68) [52] 0.83 ( 0.75) [39] 0.84 ( 0.46) [13] NS Olanzapine 5.1 ( 2.4) [27] 5.35 ( 2.4) [20] 4.8 ( 2.5) [7] NS Aripiprazole 7.8 ( 3.8) [9] 8.05 ( 3.9) [8] 5.0 [1] NS Ziprasidone 21.3 ( 26.5) [2] 21.3 (2 6.5) [2] NS Antidementia drugs Start of drug from admission, months 6.5 ( 13.0) [60] 8.6 ( 15.8) [33] 3.6 ( 6.9) [27] NS Drug, no. of subjects NS Donepezil Galantamine Memantine Rivastigmine Weight change from Admission, number NS Weight loss No change Weight gain Data expressed as mean ( SD) or median (interquartile range) and [number of subjects with analyzable data] if different from column headings. NS, not significant. * Chi square 8.72, df 1. Chi square 13.2, df 4. There were 4 subjects on memantine alone, and 7 who were taking memantine in combination with another antidementia drug. 8.1 kg/m 2, P.012). However, this was predominantly due to obesity in women with diabetes compared to those without diabetes (BMI versus kg/m 2, P.001), whereas there was no difference in BMI in men with or without diabetes (BMI versus kg/m 2, not significant). Subjects remained in the nursing homes for a median duration of 5.3 months (interquartile range ) although many remained for longer durations (maximum duration 116 months) (Figure 1). Sixty-two patients were on antipsychotic medications for indeterminate times before admission to the nursing home. Consequently, duration of antipsychotic agent use could not be determined, and follow-up analysis was performed from start of admission. During the observation period there were 9 cases of changing diabetes status, either new (n 4) or worsening of diabetes (n 5) (see Figure 2). Within 36 months of admission to the nursing home, by the life table analysis, there was a 15% incidence of the entire population of new or worsening diabetes (Figure 3). Five cases occurred after the onset of the antipsychotic medication (2 new onsets) and 4 cases (2 new onset) occurred after admission to the nursing home but before the institution of any atypical antipsychotic medication. Of those who had deterioration in their diabetes status, there were 6 women and 3 men; 7 remained weight neutral and 2 gained weight. Of those who developed a change in diabetes status there were 2 ORIGINAL STUDIES Albert et al 117

4 Number of subjects >36 Months 100% Fig. 1. Duration of observation period in the nursing home. The number of subjects followed at each time period is shown in the bars, and the cumulative percentage of the population followed is shown in the line graph. men and 2 women prescribed quetiapine, and 1 woman prescribed risperidone. None of these associations were significantly different from those who did not have a change in the diabetes status during the observation period. DISCUSSION The atypical antipsychotic agents are widely used in therapy for elderly patients with dementia who have psychosis and behavioral disturbances. In younger patients with schizophrenia, diabetes may be associated with the use of atypical antipsychotic medication, especially if there is concomitant weight gain. In the elderly, weight loss may be more of a clinical problem. In this retrospective review of elderly nursing home patients taking atypical antipsychotics, weight loss continued to be a relevant issue, and there was no temporal relationship between the use of atypical antipsychotics and deteriorating diabetes status. The Clinical Antipsychotic Trials of Intervention Effectiveness-Alzheimer s Disease (CATIE-AD) Study 12 addressed Fig. 2. Onset of DM or worsening of DM in patient population as associated with institution of atypical antipsychotic medication. 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Cumulative frequency Percent Without New Onset Diabetes 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% >36 Months Duration >36 Obervations Fig. 3. Life table analysis of subjects. Subjects who developed new or worsening diabetes during the period of observation are shown as squares for those who received the atypical antipsychotic drug before the onset of diabetes, and circles refer to those subjects who developed diabetes before the prescription of the antipyschotic medications. Error bars refer to SEM. The number of subjects under observation is shown below the graph. the short-term use of atypical antipsychotic agents in an outpatient setting of elderly individuals with Alzheimer s dementia. Our study population was similar to the CATIE-AD study in the degree of obesity, as calculated by BMIs, between the 2 studies (BMI versus kg/m 2, respectively, not significant). The doses of medications were similar for this study and the CATIE-AD for quetiapine (59.3 mg versus 56.5 mg), risperidone (0.8 mg versus 1.0 mg), and olanzapine (5.1 mg versus 5.5 mg per day). However, the population evaluated in this study differed from the CATIE-AD in many other ways. This was a nursing home population, whereas CATIE-AD was essentially an outpatient population. Subjects in this study were older than in CATIE-AD ( years versus years, P.006), there was a greater percentage of women (65.5% versus 55%, P.036), and the subjects were predominantly white (94% versus 81%). Subjects were maintained on the atypical antipsychotics for a median duration of 13.1 weeks in this study, and for a median duration of 5.3 to 8.1 weeks in CATIE-AD. In the CATIE-AD study there was weight gain (defined as 7% increase) in 10% of those on atypical antipsychotics. In our study, using the minimum data set (MDS) criteria, which records weight change of plus or minus 5%, 13% had weight gain. The mean admission BMI of 24.5 kg/m 2 is considered within normal weight distribution. Nine subjects either developed new-onset diabetes (n 4) or worsening of existing diabetes (n 5) during the observation period. Over a 36-month observation period, approximately 15% had worsening of diabetes, similar to the incidence rate reported in the literature of 4.5 cases per year in this age range. 14 This worsening of diabetes occurred in 5 subjects after starting the antipsychotic drugs, and in 4 subjects the diabetes antedated the prescription of the antipsychotic drugs, and so it is difficult to assign causality for drug use to deterioration of diabetes status. 118 Albert et al JAMDA February 2009

5 The study described here addresses the clinical practice of using atypical antipsychotic agents in a nursing home population. In this nursing home setting, there were a greater percentage of women than men (66% versus 34%), and a greater proportion of the oldest-old ( 85 years of age) women compared with men (57% versus 40%). In younger patients on atypical antipsychotics, weight gain may be associated with metabolic deterioration in glucose and lipids. 7 9 However, in this population weight loss occurred in 32%, rather than weight gain, which occurred in 13%. We did not study a control population not on these drugs and so it is unknown whether the proportion of weight changes may have been affected by these medications. In general, weight loss is considered more of a concern in this elderly population. If it could be demonstrated that these atypical antipsychotics do not cause metabolic derangements, then drug-induced weight gain may not be considered an impediment in this elderly population. Another limitation was the short duration of drug therapy of 3.1 months (interquartile range months), although that was the clinically effective duration of therapy in this nursing home population, and comparable to the described duration of follow-up in the CATIE-AD study. 12 CONCLUSIONS In an elderly long-term care NH population, there was no evidence that the short-term use (median 13.1 weeks) of atypical antipsychotic agents was associated with the onset or worsening of diabetes mellitus. REFERENCES 1. Sink KM, Holden KF, Yaffe K. Pharmacological treatment of neuropsychiatric symptoms of dementia. A Review of the Evidence JAMA 2005; 293: Schneider LS, Dagerman KS, Insel P. Risk of death with atypical antipsychotic drug treatment for dementia. Meta-analysis of randomized placebo-controlled trials. JAMA 2005;294: Wang PS, Schneeweis S, Avorn J, et al. Risk of death in elderly users of conventional vs. atypical antipsychotic medications. N Eng J Med 2005; 353: Schneider LS, Tariot PN, Lyketsos CG, et al. National Institute of Mental Health Clinical Antipsychotic Trials of Intervention Effectiveness (CATIE) Alzheimer Disease Trial Methodology. Am J Geriatr Psychiatry 2001;9: American Diabetes Association. American Psychiatric Association. American Association of Clinical Endocrinologist. North American Association for the Study of Obesity. Consensus Development Conference on Antipsychotic Drugs and Obesity and Diabetes. Diabetes Care 2004;27: Gianfrancesco FD, Grogg AL, Mahmoud RA, et al. Differential effects of risperidone, olanzapine, clozapine, and conventional antipsychotics on type 2 diabetes: Findings from a large health plan database. J Clin Psychiatry 2002;63: Buse JB, Cavazzoni P, Hornbuckle K, et al. A retrospective cohort study of diabetes mellitus and antipsychotic treatment in the United States. J Clin Epidemiol 2003;56: Leslie DL, Rosenheck RA. Incidence of newly diagnosed diabetes attributable to atypical antipsyhcotic medications. Am J Psychiatry 2004;161: Newcomer JW. Abnormalities of glucose metabolism associated with atypical antipsychotic drugs. J Clin Psychiatry 2004;65(suppl): Ader M, Kim SP, Catalano KJ, et al. Metabolic dysregulation with atypical antipsychotics occurs in the absence of underlying disease: a placebo-controlled study of olanzapine and risperidone in dogs. Diabetes 2005;54: Henderson DC, Cagliero E, Copeland PM, et al. Glucose metabolism in patients with schizophrenia treated with atypical antipsychotic agents: A frequently sampled intravenous glucose tolerance test and minimal model analysis. Arch Gen Psychiatry 2005;62: Schneider LS, Tariot PN, Dagerman KS, et al for the CATIE-AD study group. Effectiveness of atypical antipsychotic drugs in patients with Alzheimer s disease. N Eng J Med 2006;355: Colton T. Longitudinal studies and the use of the life table. In: Statistics in Medicine. Boston: Little Brown and Co, pp Kenney SJ, Aubert RE, Geiss LS. Prevalence and incidence of noninsulin-dependent diabetes mellitus. In: Harris MI, Cowie CC, Stern MP, et al, eds. Diabetes in America. 2nd edition. Bethesda, MD: National Institutes of Health, pp ORIGINAL STUDIES Albert et al 119